Unusual periodic modulation in the radio emission of the methane dwarf binary WISEP J101905.63+652954.2

TL;DR

This study presents a six-year LOFAR monitoring campaign of the methane-dwarf binary J1019+65, revealing persistent, highly circular ECMI radio emission modulated by rotation with a known period $P_1\approx3.07$ h and a newly identified shorter period $P_2\approx0.787$ h. Through cross-epoch Lomb-Scargle analysis, the authors identify two significant frequencies, $f_1\approx0.325\,\mathrm{h^{-1}}$ and $f_2\approx1.271\,\mathrm{h^{-1}}$, and conduct extensive fidelity tests to rule out artefacts, while phase-folded light curves and a stacked image confirm the emission origin and geometry consistent with a Jupiter-like engine. They evaluate four scenarios for $f_2$—aliasing, harmonics, the rotation of the second brown dwarf, or satellite interaction—and conclude that aliasing and harmonics are unlikely, with rotation of the second BD remaining plausible but requiring infrared confirmation; tidal-satellite scenarios are disfavoured by dynamical constraints. By combining Poisson and binomial statistics with LOFAR non-detections, they constrain the average duty cycle of a radio-loud brown dwarf to $\langle D \rangle_{\rm BD}=0.030^{+0.034}_{-0.030}$ and the observed radio-loud fraction to $F'_{\rm radio}=0.088^{+0.168}_{-0.088}$, implying a beaming fraction $F_{\rm beam}\approx0.127$ and an intrinsic radio-loud fraction $F_{\rm radio}\lesssim0.691$; thus such objects are not extremely rare but exhibit low duty cycles. The results underscore the need for targeted infrared monitoring to confirm the second period and for broader, long-term surveys to refine population statistics of ECMI in ultracool dwarfs and their potential exoplanetary companions.

Abstract

Brown dwarfs display Jupiter-like auroral phenomena, such as rotationally modulated electron cyclotron maser radio emission. Radio observations of cyclotron maser emission can be used to measure their magnetic field strength, topology, and to deduce the presence of magnetically interacting exoplanets. Observations of the coldest brown dwarfs (spectral types T and Y) are especially intriguing, as their magnetospheric phenomena could closely resemble those of gas-giant exoplanets. Here we report observations made over ten epochs, amounting to 44 hours, of WISEP J101905.63+652954.2 (J1019+65 hereinafter) using the LOFAR telescope between 120 and 168 MHz. J1019+65 is a methane dwarf binary (T5.5+T7) whose radio emission was originally detected in a single-epoch LOFAR observation to be highly circular polarised and rotationally modulated at $\approx 3$h. Unexpectedly, our long-term monitoring reveals an additional periodic signature at $\approx 0.787$h. We consider several explanations for the second period and suggest that it could be the rotationally modulated emission of the second brown dwarf in the binary, although follow-up infrared observations are necessary to confirm this hypothesis. In addition, the data also allow us to statistically estimate the duty cycle and observed radio-loud fraction of the 120-168\,MHz cyclotron emission from methane dwarfs to be $\langle D \rangle = 0.030^{+0.034}_{-0.030}$ and $F^{'}_{\rm radio} = 0.088^{+0.168}_{-0.088}$ respectively.

Figures (16)

• Figure 1: LOFAR observation of J1019+65 carried out on 2021-02-33. Top panel: Stokes V light curves with a cadence of 4 minutes at different frequencies. The shaded region represents $\pm 1 \sigma$ uncertainty. The light curve at 152--168 MHz is omitted for the sake of clarity. The top axis represents the time of observation in Modified Julian Dates (MJD). The black dashed line at 0 mJy is drawn for clarity. Bottom panel: Lomb-Scargle periodogram of the radio light curves. The three red dashed lines represent the necessary LS power (i.e. peak height) to achieve a false-alarm probability (FAP) of 5, 1, and 0.3. The grey curve represents the (negative) LS periodogram of the window function, which is a light curve with the same timestamps as the original curve, but the flux density of each data point is replaced with unity (i.e. a flat light curve). The two relevant frequencies $f_1 = 0.315 \pm 0.013\,\unit{h^{-1}}$ and $f_2 = 1.247 \pm 0.013\,\unit{h^{-1}}$ are indicated by the black arrows in the Lomb-Scargle periodogram. The 120--136 MHz (orange) peak at $f_1$ is consistent with the original frequency discovered by harish_j1019, while the relevancy of the $f_2$ is discussed in Sects. \ref{['sec:results-fidelity']}. The frequency uncertainties were computed using Eq. 52 by vanderplas2018.
• Figure 2: Lomb-Scargle periodograms of the LOFAR cross-epoch radio light curve, which includes data from both the original LoTSS DR2 epoch and our 10 LOFAR follow-up epochs. The spectrum reveals the peak at $f_1 \approx 0.325\,\unit{h^{-1}}$ originally discovered by harish_j1019. Another significant peak not previously discovered is detected at $f_2 \approx 1.271\,\unit{h^{-1}}$, with a false alarm probability (FAP) of less than 0.3. The locations of these two peaks are indicated by the black arrows. The three red dashed lines represent the FAP of different values. The grey curve represents the LS periodogram of the window function, which is a light curve with the same timestamps as the original curve, but the flux density of each data point is replaced by unity (i.e. a flat light curve). The inset plot in each panel shows a zoomed-in version of the observed J1019+65 power spectrum in the vicinity of the tallest LS peak, with the same three FAP red dashed lines.
• Figure 3: Histogram of peak LS power calculated from light curves at random sky locations in the vicinity of J1019+65.
• Figure 4: Histogram of peak LS power calculated by randomly shuffling the measured light curve values while keeping the same time sampling window.
• Figure 5: Stokes V cross-epoch (2017--2023) light curve of J1019+65 phase-folded at a period of $P_1=3.0742$ h corresponding to the $f_1$ peak seen in Fig. \ref{['fig:cross-epoch-ls']}. The light curve is binned at intervals of 0.04 in phase and one full period corresponds to a phase interval of unity. The choice of starting phase is arbitrary. The green line represents the best sinusoidal fit using the computed Lomb-Scargle model. The J1019+65 radio burst shown in Fig. \ref{['fig-app:lofar-2017-multi-lcls']} is also plotted as a red line (with its width representing the duration of the burst) to indicate at which phase relative to the phase-folded light curves does it correspond to. Note that the burst shown in red only occurs in one epoch out of the whole dataset. The black dashed line at 0 mJy is drawn for clarity.